Discovery of Conserved RNA Secondary Structures in Farm Animals

The molecular structure of especially long RNA transcripts is still largely unexplored, but often directly associated with their functionality. For instance, structured RNA domains are involved in binding nucleic acids and proteins as well as RNA stabilization. Recent transcriptome-wide in-vitro structure probing experiments have revealed the complexity of the structure landscape; however, these attempts do not consider that functional important RNA domains will be under phylogenetic constraints.

We present a large collection of RNA secondary structures conserved in vertebrates (CRSs) that we computationally predicted in a genome-wide screen using human as the reference organism. CRSs overlap a range of known sequences, from Rfam to long non-coding RNAs, and with most of the CRSs in intronic or intergenic regions. Many of the RNA structures are under purifying selection, despite of that many have low sequence identity (<50%). Comparing the CRSs to causal autoimmune disease variants, we predict several SNPs to cause changes in the RNA structure. CRSs are significantly co-localized with RNA binding sites of many proteins and cis-regulatory regions. The latter comprise enhancer and promoter regions which have recently been shown to be transcribed as eRNAs and PROMPTs. A large fraction of the CRSs are located in transcribed genomic regions, and we show evidence that their expression is conserved in human and mouse by analyzing total RNAseq and qPCR. Finally, we use conservation on the level of RNA structure, rather than primary sequence, to discover structured RNAs in farm animals.